基于协方差矩阵和导向矢量不确定性的鲁棒波束形成技术

发布时间：2018-10-14 09:03

【摘要】：自适应波束形成技术是阵列信号处理中一个重要的研究方向,其在通信、雷达、声呐、语音处理、医学成像等领域,都有着广阔的应用前景。传统的波束形成方法旨在保持期望信号一定的前提下抑制干扰。然而,在现实应用中,由于环境、信号源以及天线阵元的真实情况与预设情况存在偏差,传统波束形成方法的性能会因这些偏差而急剧下降。因此,研究如何提高鲁棒性能对自适应波束形成方法而言意义重大。近几十年,研究人员提出了许多具有鲁棒性的自适应波束形成方法,对角加载是其中一种重要的方法。该方法在采样数少于阵列传感器数目导致样本协方差矩阵不可逆时仍能获得较好的性能。本论文在对角加载范畴内,提出了一种新的鲁棒波束形成方法,主要工作内容如下:(1)介绍了波束形成的相关基础知识,介绍了几种常见的波束形成方法,给出了其主要推导过程,并结合推导过程对其特点和缺陷进行了分析说明。(2)详细介绍了几种重要的对角加载范畴内的鲁棒自适应波束形成方法,介绍了其研究状况,推导过程,计算复杂度,并对其性能优劣点进行了比较分析。(3)提出了一种新的可变对角加载鲁棒自适应波束形成方法。通过研究分析如最差性能优化和RCB等考虑导向矢量不确定性集的鲁棒方法,发现这些对角加载范畴内的方法都只单独考虑了导向矢量的不确定性。因此,本文所提方法的一个创新点在于,假设协方差矩阵和导向矢量都存在误差,并对这两种误差分别施加约束条件,进而最大化SINR。根据约束条件进行推导,本文得到一个最小最大优化问题,根据最优化的相关结论,本文对模型进行放缩,从而得到一个最大最小优化问题。最终,通过KKT优化条件求解出这个最大最小优化问题。此外,由于对角加载方法其作用的实质是通过人工投影白噪声以降低输入信噪比,导致大的加载量虽可以增强鲁棒性能却使系统对干扰和噪声的抑制能力下降。因此,本论文的另一个创新点在于,所提方法算出的可变对角加载量考虑了特征值的大小,即大特征值对应相对小的加载量,而小特征值对应相对大的加载量。(4)给出了本文所提方法与前述几种对角加载方法的仿真结果,并对仿真结果进行了分析比较,说明了所提方法性能更优。
[Abstract]:Adaptive beamforming technology is an important research direction in array signal processing. It has a broad application prospect in communication, radar, sonar, speech processing, medical imaging and other fields. The traditional beamforming method is designed to suppress interference on the premise of keeping the desired signal. However, in practical applications, the performance of traditional beamforming methods will be drastically reduced due to the deviation between the environment, signal source and antenna array elements and the preset conditions. Therefore, it is important to study how to improve robust performance for adaptive beamforming. In recent decades, many robust adaptive beamforming methods have been proposed, among which diagonal loading is one of the most important methods. The proposed method can still achieve good performance when the number of samples is less than the number of array sensors and the sample covariance matrix is irreversible. In this paper, a new robust beamforming method is proposed under diagonal loading. The main work is as follows: (1) the basic knowledge of beamforming is introduced, and several common beamforming methods are introduced. The main derivation process is given, and its characteristics and defects are analyzed. (2) several important robust adaptive beamforming methods in diagonal loading category are introduced in detail, and their research status and derivation process are introduced. The computational complexity and performance advantages and disadvantages are compared and analyzed. (3) A new robust adaptive beamforming method with variable diagonal loading is proposed. Based on the analysis of robust methods such as worst performance optimization and RCB, it is found that these methods in diagonally loaded category only consider the uncertainty of guidance vector. Therefore, one of the innovations of the method proposed in this paper is to assume that there are errors in both the covariance matrix and the guidance vector, and to impose constraints on the two errors to maximize the SINR.. According to the constraint conditions, this paper obtains a minimum and maximum optimization problem. According to the relevant conclusions of optimization, the model is scaled down, and a maximum and minimum optimization problem is obtained. Finally, the maximum and minimum optimization problem is solved by KKT optimization condition. In addition, the essence of diagonal loading method is to reduce the input SNR by artificial projection of white noise, which results in a large amount of loading can enhance the robust performance but reduce the ability of the system to suppress interference and noise. Therefore, another innovation of this paper is that the variable diagonal load calculated by the proposed method takes into account the magnitude of the eigenvalue, that is, the large eigenvalue corresponds to a relatively small amount of loading. The small eigenvalues correspond to a relatively large amount of loading. (4) the simulation results of the proposed method and the above diagonal loading methods are given, and the simulation results are analyzed and compared, which shows that the proposed method has better performance.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.7

【相似文献】